L’intelligenza artificiale alla prova del diritto europeo: verso il diritto della paura?

Abstract

In the list of "five great objectives" established in art. 3, the Treaty on European Union (TEU) includes "scientific and technological progress" among the measures to be promoted in order to achieve a free, inclusive and competitive internal market. Our Constitution establishes that public administrations must ensure "the balance of budgets and the sustainability of public debt" (art. 97, paragraph 1, Constitution). It then identifies (in art. 97, paragraph 2) a legal reservation on the matter ("Public offices are organized according to legal provisions....") and also indicates the standards of conduct both for the apparatus and for the activities of thethemselves carried out (“...so that the good performance and impartiality of the administrationare ensured”). Although the theme of "technological development" is absent in the Constitution,obviously including that linked to AI, as a behavioral canon of the public administration, no one doubts that the public sector must use modern technologies, first and foremost AI (Avanzini, 2019). In particular, despite the absence of a regulatory framework, the administrative judge has no doubts about it, who has long considered the use of AI by the public administration to be constitutionally compatible.